Target location based on quasi-Newton algorithm in multistatic passive radar

The solution to nonlinear equations is required in passive radar positioning. However, when the traditional Newton method is adopted, Newton direction may not exist since Jacobian matrix of the equations may be singular. Although the quasi-Newton algorithm overcomes the singular problems, it confronts the common problem of converging to a local optimum value instead of the global optimum value. In this paper, we introduce two quasi-Newton algorithms, namely Spectral Approach for Nonlinear Equations (SANE) and Nonmonotone Spectral for Nonlinear Equations (NMSGNE), to solve this problem. The effectiveness of the proposed target location algorithms is verified via simulation.